The previous three year grant R01MH42643 funded (a) the ascertainment of 10 MDI pedigrees and (b) laboratory equipment to initiate genotyping. With support we ascertained 10 moderate size multiplex MDI pedigrees demonstrating unilineal transmission of illness. In addition to meeting the primary goals of the original grant, we carried out simulation runs to evaluate the power of our sample for linkage analysis, and genotyped 8 of 10 pedigrees with over 328 DNA markers. Our manuscript highlighting our genotyping effort - "A Genome Wide Search for Genes Predisposing to Manic-Depression Assuming Autosomal Dominant Inheritance" - is in press in the American Journal of Human Genetics. The primary aim of the proposed grant renewal is to identify genes that predispose to manic-depression in a subgroup of families. This will be carried out by: 1. Linkage analysis techniques using 13 moderately-sized multiplex families. Simulation analyses using 13 families already identified assuming genetic homogeneity and 5 to 23% phenocopies among the affected cases, indicate that we have 100% power to detect linkage with a polymorphic DNA marker mapping 2.5% away from the disease gene. Moreover, given 30% unlinked pedigrees, we have 82% power to detect linkage. 2. These pedigrees will be genotyped using: a. Polymorphic candidate genes which may play a role in the pathophysiology of MDI, such as serotoninergic and catecholaminergic receptor subtype genes. b. A complete linkage map with DNA markers spaced at 5 centimorgan intervals. The genotyping effort is focused on highly polymorphic simple sequence repeat markers. 3. The genetic data will be analyzed using the lod score as well as the affected pedigree member method. The primary hypothesis is that MDI is transmitted by a dominant allele with reduced penetrance and variable expressivity, but the data will also be analyzed using the affected pedigree member method, a nonparametric linkage test in which data are analyzed without specifying mode of inheritance. 4. Mutation scanning techniques in candidate genes that map to a promising area of linkage. We will use simulation analyses to evaluate promising lod scores for the total sample of pedigrees as well as individual families. If a polymorphic candidate gene shows significant linkage in two or more bipolar families, and if the genomic sequence is available we will screen for mutations in the coding regions of the candidate gene by using single strand conformation polymorphism (SSCP) analysis.